Diagnostic accuracy of the Xpert® MTB/RIF cycle threshold level to predict smear positivity: a meta-analysis.

SETTING: Xpert® MTB/RIF is the most widely used molecular assay for rapid diagnosis of tuberculosis (TB). The number of polymerase chain reaction cycles after which detectable product is generated (cycle threshold value, CT) correlates with the bacillary burden.OBJECTIVE To investigate the association between Xpert CT values and smear status through a systematic review and individual-level data meta-analysis. DESIGN: Studies on the association between CT values and smear status were included in a descriptive systematic review. Authors of studies including smear, culture and Xpert results were asked for individual-level data, and receiver operating characteristic curves were calculated. RESULTS: Of 918 citations, 10 were included in the descriptive systematic review. Fifteen data sets from studies potentially relevant for individual-level data meta-analysis provided individual-level data (7511 samples from 4447 patients); 1212 patients had positive Xpert results for at least one respiratory sample (1859 samples overall). ROC analysis revealed an area under the curve (AUC) of 0.85 (95%CI 0.82-0.87). Cut-off CT values of 27.7 and 31.8 yielded sensitivities of 85% (95%CI 83-87) and 95% (95%CI 94-96) and specificities of 67% (95%CI 66-77) and 35% (95%CI 30-41) for smear-positive samples. CONCLUSION: Xpert CT values and smear status were strongly associated. However, diagnostic accuracy at set cut-off CT values of 27.7 or 31.8 would not replace smear microscopy. How CT values compare with smear microscopy in predicting infectiousness remains to be seen.

Performance of the Xpert® MTB/RIF assay in an active case-finding strategy: a pilot study from Tanzania.

In this pilot study, we evaluated the Xpert® MTB/RIF assay in an active case-finding strategy, using two spot sputum samples collected within a 1-hour interval from household contacts of smear-positive TB index cases. Tuberculosis (TB) confirmed by culture served as the reference standard. Among 219 enrolled contacts, the yield of active TB was 2.3%. While the sensitivity of smear microscopy was 60% (95%CI 14.7-94.7), Xpert MTB/RIF achieved a sensitivity of 100% (95%CI 47.81-100.0). All culture-confirmed cases tested positive by Xpert MTB/RIF on the first submitted sample, suggesting that the evaluation of only one sample could be sufficient for TB diagnosis in this context.